Latest AI Insights

Model Evaluations

Model Directory

Model Comparison

Resource Center

Search blog

加载中...

DataLearner AI

A knowledge platform focused on LLM benchmarking, datasets, and practical instruction with continuously updated capability maps.

产品

Leaderboards
模型对比
Datasets

资源

Tutorials
Editorial
Tool directory

关于

关于我们
隐私政策
数据收集方法
联系我们

© 2026 DataLearner AI. DataLearner curates industry data and case studies so researchers, enterprises, and developers can rely on trustworthy intelligence.

隐私政策服务条款

Qwen1.5-MoE-A2.7B - AI Model Details and Benchmarks | DataLearnerAI

Page navigation

目录

Model catalogQwen1.5-MoE-A2.7B

QW

Qwen1.5-MoE-A2.7B

Qwen1.5-MoE-A2.7B

Release date: 2024-03-28更新于: 2024-04-07 22:05:35501

Live demoGitHub Hugging Face Compare

Parameters

143.0亿

Context length

1K

Chinese support

Supported

Reasoning ability

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Qwen1.5-MoE-A2.7B

Model basics

Reasoning traces

Not supported

Context length

1K tokens

Max output length

No data

Model type

基础大模型

Release date

2024-03-28

Model file size

Qwen1.5-MoE-A2.7B

Open source & experience

Code license

Tongyi Qianwen RESEARCH LICENSE AGREEMENT

Weights license

Tongyi Qianwen RESEARCH LICENSE AGREEMENT- 免费商用授权

GitHub repo

https://github.com/QwenLM/Qwen1.5

Hugging Face

Qwen1.5-MoE-A2.7B

Official resources

Paper

Qwen1.5-MoE: Matching 7B Model Performance with 1/3 Activated Parameters

DataLearnerAI blog

重磅！阿里巴巴开源自家首个MoE技术大模型：Qwen1.5-MoE-A2.7B，性能约等于70亿参数规模的大模型Mistral-7B

Qwen1.5-MoE-A2.7B

API details

API speed

No data

No public API pricing yet.

Qwen1.5-MoE-A2.7B

Benchmark Results

No benchmark data to show.

Qwen1.5-MoE-A2.7B

Publisher

阿里巴巴

View publisher details

Qwen1.5-MoE-A2.7B

Model Overview

阿里巴巴开源的阿里系首个MoE大语言模型，143亿参数模型效果与70亿差不多，但是每次推理只激活27亿参数，可以在手机端运行。

DataLearner 官方微信

欢迎关注 DataLearner 官方微信，获得最新 AI 技术推送

DataLearner 官方微信二维码

3.46GB

MoE architecture

No

Total params / Active params

143.0B / N/A

Knowledge cutoff

No data

Inference modes

No mode data

https://huggingface.co/Qwen

Live demo

No live demo